CPU usage in lxc container

[g0ha]

Hi! Some exotic problems arriverd
Recently i move lxc container from host with pve 4.2-18 to host with pve 6.0-11. Just created new container, then move .raw file from old host to new. I moved 4 containers, 3 works fine, but 1 displays 100% cpu usage by host utils like top or htop. Before moving cpu usage was about 1-2%. First strange thing: that PVE dashboard do not see that usage, it reports normal cpu load. Second: in this pve 6.0-11 host already was 1 container and after moving (those new 4) he began to behave the same as this, cpu usage 100%, PVE dashboard reports about normal load.
Looks like some counters are broken, where to dig?

 

[g0ha]

top -i:
top - 16:21:13 up 17:28, 0 users, load average: 3.93, 4.90, 4.23
Tasks: 43 total, 1 running, 42 sleeping, 0 stopped, 0 zombie
%Cpu(s):100.0 us, 0.0 sy, 0.0 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
KiB Mem: 2097152 total, 373412 used, 1723740 free, 0 buffers
KiB Swap: 2621440 total, 6648 used, 2614792 free. 171196 cached Mem

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
9280 user 20 0 355776 43020 24660 S 0.7 2.1 7:36.42 php5-fpm
2034 zabbix 20 0 95152 2828 1972 S 0.3 0.1 0:13.27 zabbix_agentd

In attach vm's dashboard